In this application we propose several studies to substantially improve our understanding of the role of fatty acids, from both diet and endogenous production, in breast carcinogenesis. Dietary fat has long been hypothesized to increase breast cancer risk, but no overall association has been observed in most large prospective studies. Despite the limitations of dietary fat intake as an exposure, associations with specific types of fat (e.g., animal and trans fats) and/or in specific subsets of women (e.g., premenopausal) have been observed in some studies. Biomarkers are likely to reflect the internal dose of dietary fatty acids more closely and allow for an examination of endogenously synthesized fatty acids, which make up a substantial component of internal exposure to fatty acids. We propose to measure erythrocyte fatty acids in a nested case-control study within the Nurses' Health Study II (NHSII), an established cohort of younger women. First, we will assess the association of saturated and trans fatty acids that reflect intake of dairy fat and meat (15:0, 16:1n- 7t, 18:1n-7t) and trans fatty acids from partially hydrogenated oils (18:1 and 18:2) with breast cancer risk. We also will assess the association with n-3 and n-6 polyunsaturated fatty acids. To further explore possible mechanisms, we will examine the associations of these specific fatty acids with tumor tissue over-expression of inflammatory markers. Next, we will address an important component of endogenously synthesized fatty acids, activity of 9-desaturase (a.k.a. stearoyl-CoA desaturase 1 (SCD1)), measured by the ratio of specific saturated to monounsaturated fatty acids, (saturation index (SI)n-7 and SIn-9). Experimental evidence suggests SCD1 is important in breast tumor growth and the limited epidemiologic evidence to date suggests an increased breast cancer risk with higher SCD1 activity. We will further investigate this association with tumor over-expression of fatty acid synthase (FAS). We will utilize archived (stored at d -130C) blood samples collected in 1996-1999 from 29,611 women in the NHSII. We will conduct a nested case-control study using these blood samples and archived tumor tissue blocks from approximately 75% of cases. We anticipate more than 900 incident cases of breast cancer, including nearly 500 women premenopausal at diagnosis and more than 200 women who were premenopausal at blood collection but postmenopausal at diagnosis. We expect approximately 600 cases with tumor tissue. This study provides an ideal context in which to investigate these associations, given the prospective nature, extensive covariate information, archived blood and tissue specimens, and an ongoing, high rate of follow-up. Our proposed study will be the first detailed assessment of these associations among premenopausal women and should contribute substantially to our understanding of the associations of fatty acids with breast cancer risk, potentially leading to important recommendations for breast cancer prevention as well as new avenues of exploration in the SCD1 and FAS pathways.